1. Field of the Invention
The present invention relates to an image processing apparatus capable of performing error diffusion processing and ordered dither gradation processing.
2. Description of the Related Art
In general, in a document image processing apparatus capable of handling not only code information but also image information, an original image read by a reading means such as a scanner is performed such that simple binarization (or digitization) processing is performed with respect to image information having contrast, e.g., character and line graphic information, by using fixed threshold values, while binarization processing based on a pseudo-gradation means such as an ordered dither method is performed with respect to image information having gradation, e.g., photographic information. With this operation, a print including a portion with gradation is output from a binary printer or the like which has a small number of available gradation levels.
An ordered dither binarization method (reference: Design of dither waveforms for quantized visual signals by Limb J. O., Bell System Tech. J. 48, pp. 2555-2582, 1969) is a method of performing binarization processing by adding a periodic dither signal to an input pixel or a threshold value.
In multi-gradation level processing performed by using a conventional ordered dither method, when the number of output gradation levels of data obtained after processing is small, dither threshold values can be generated by using a plurality of memories or performing a memory read operation a plurality of number of times. However, as the number of gradation levels increases, a memory and a register for storing a larger number of dither threshold values are required.
In addition, if a memory device such as an SRAM is used as a memory for storing dither threshold values in accordance with the size of a dither matrix, a plurality of memories corresponding to the required number of gradation levels are prepared, or one memory in which all threshold values are stored is prepared, and threshold values corresponding to the number of gradation levels are read out from the memory or memories in units of pixels. If a plurality of memories are prepared, the number of data lines increases, resulting in a complicated structure. If one memory is prepared, since read operations corresponding to the number of gradation levels are performed in units of pixels, a long processing time is required, resulting in a decrease in gradation processing speed.